USRA · SaaS Systems
Practical systems for people shipping SaaS
Infrastructure, rate limits, auth, automation, and production patterns — written operator-to-operator, with real numbers and real failure modes.
New playbooks and articles every week
InfrastructureRate LimitsAuthWebhooksBillingAutomation
Built for the person who ships it
One to five people, a live product or one close to launch, and no platform team to absorb operational mistakes. Every article assumes you write the code, run the deploy, and answer the support email.
01 / Patterns
Production patterns
The concrete shape of a solution, including what it costs you in complexity. Nothing gets written up until it has run against real traffic.
02 / Checklists
Pre-flight checklists
The short list to verify before a system touches money, identity, or third-party quota — the three places early SaaS loses value quietly.
03 / Postmortems
Real failure stories
What broke, the symptom that surfaced first, the wrong fix, and the one that held. Specific numbers instead of general advice.
Start with these
Three articles that cover the decisions most early SaaS systems get wrong in the first month of real traffic.
Infrastructure
The One-Server Question
What to run yourself before you have customers — and the three signals that justify splitting the stack.
API Workflows
Webhook Idempotency
The five-line habit that turns guaranteed duplicate delivery into a harmless no-op instead of a double charge.
Automation
Budget Your Outbound Calls
Size a shared token bucket under the documented limit so 429s become a safety net, not your throttling mechanism.
Free playbooks
Assets rather than essays: checklists, decision frameworks, and ordered playbooks you can apply the same day. All free, no signup required.
01 / Framework
The Early-SaaS Stack Decision Checklist
Hosting, database, auth, payments, email — a recommended default for each layer and the condition that should make you deviate.
02 / Setup guide
Auth in a Weekend
Magic links, 15-minute tokens, a sessions table with real revocation, and rotation with reuse detection. Four tables total.
03 / Checklist
Billing Webhooks Done Right
Idempotency on the event id, an append-only ledger, ordering guards, and replay as a one-command operation.
04 / Playbook
Rate-Limit Playbook
What to limit first, in order, with the bucket sizes and thresholds we actually run in production.
05 / Checklist
Minimal Observability
The five alerts worth waking up for, their thresholds, and why everything else is noise at this stage.
06 / Playbook
The First-Paying-Customer Playbook
Four stages with an exit condition each, ordered so the cheapest information arrives first.
Get new articles
Practical systems for shipping SaaS, straight to your inbox. No spam.
One email per new article. Unsubscribe any time.
What this is not
No course, no cohort
There is nothing to enroll in. The articles are free and stay free.
No framework worship
Recommendations are scoped to a stage and a constraint, and say when to stop using them.
No theory without traffic
If a pattern has not survived production, it does not get published here.